International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 11 Issue: 08| Aug2024 www.irjet.net p-ISSN:2395-0072
Harvesting Knowledge: A Critical Review of Ontology-Based Information Extraction Across Diverse Scientific Domains
Christeena Raphy1 , Aparna Mohan2
1 M.Tech Student, Dept. of Computer Science, Malabar College of Engineering and Technology, Kerala, India
2 Assistant Professor, Dept. of Computer Science, Malabar College of Engineering and Technology, Kerala, India
Abstract - Ontology-Based Information Extraction (OBIE) has gained prominence as a pivotal technique for transforming unstructured data into structured knowledge, leveraging the semantic and organizational strengths of ontologies. This critical review explores the diverse applications and methodologies of OBIE across various scientific domains, emphasizing its role in enhancing data interoperabilityandprecision.Keyadvancementsincludethe integration with the Semantic Web, the development of expressive ontologies, and the implementation of machine learning and deep learning techniques. The literature highlights the versatility of OBIE, with applications ranging from academic knowledge repositories to monitoring online contentandgeologicaldataexploration.Comparativeanalysis reveals the strengths and limitations of various OBIE methodologies, such as rule-based approaches, machine learning,semantic-basedontologymapping,andtransformerbased learning. While rule-based methods are simple to implement, machine learning and advanced NLP techniques offer scalability and higher precision but require significant computational resources. The review suggests that future research should focus on hybrid models that combine these methodologies to enhance scalability, real-time processing, and interoperability. This comprehensive synthesis not only underscoresthetransformativepotentialofOBIEinscientific researchbut alsosets thestageforfurther innovations inthe field.
Key Words: Data Interoperability, Information Extraction, Machine Learning , Ontology-Based Information Extraction (OBIE), Semantic Web, Structured Knowledge
1.INTRODUCTION
Intheeraofbigdata,thechallengeofefficientlyextracting andstructuringvastamountsofinformationfrom diverse and often unstructured datasets is paramount. OntologyBased Information Extraction (OBIE) has emerged as a powerfulmethodtoaddressthischallengebyleveragingthe semantic richness and organizational capabilities of ontologies.OBIEenablesthesemanticinterpretationofdata, thereby enhancing the precision and relevance of information extraction processes across various scientific domains.
ThesignificanceofOBIEliesinitsabilitytotransformraw data into structured knowledge, which is essential for advancingresearchandinnovation.Byutilizingontologies, OBIEsystemscanprovideamoreaccurateandmeaningful representation of data, facilitating better interoperability and more sophisticated querying capabilities. This is particularly important in scientific research, where the abilitytointegrateandanalyzedatafrommultiplesourcesis crucial for making informed decisions and driving discoveries.
2. LITERATURE REVIEW
Ontology-based information extraction (OBIE) has emergedasapivotalmethodforharnessingandstructuring knowledgefromvastanddiversedatasetsacrossscientific domains.Theutilizationofontologiesfacilitatesthesemantic interpretationofdata,enhancingtheprecisionandrelevance of information extraction processes. Thisliteraturereview synthesizesrecentadvancementsandmethodologiesinOBIE, focusing on the integration with the Semantic Web, the developmentofexpressiveontologies,andtheapplicationof OBIEinvariousscientificcontexts.
Martinez-Rodriguez et al. [1] provide a comprehensive surveyontheintersectionofinformationextraction(IE)and theSemanticWeb.Theydiscusstheevolutionoftechniques andtoolsthatleveragesemantictechnologiestoenhanceIE processes.Theirworkhighlightsthecriticalroleofontologies in improving data interoperability and enabling more sophisticated semantic queries. This foundational understanding sets the stage for exploring specific applicationsandmethodologiesinOBIE.
Petrucci emphasizes the importance of learning expressive ontologies for the Semantic Web. He presents methodologiesthatenhancetheexpressivenessofontologies, enablingmorenuancedandaccurateinformationextraction [2]. His work underlines the challenges of balancing complexityandusabilityinontologydevelopment,whichis crucialforeffectiveOBIE.
Suganya and Porkodi provide a review of OBIE techniques,highlightingthebenefits ofusingontologiesto improvetheaccuracyandefficiencyofinformationextraction [3].Theirreviewcoversvariousalgorithmsandframeworks
International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 11 Issue: 08| Aug2024 www.irjet.net p-ISSN:2395-0072
that have been developed to leverage ontologies for structureddataextractionfromunstructuredsources.
Jose et al. [4] propose an OBIE framework specifically designed for academic knowledge repositories. Their framework aims to automate the extraction of academic knowledge, enhancing the accessibility and usability of scholarly information. This application demonstrates the potentialofOBIEtotransformacademicdatamanagement andretrieval.
Zaman et al. [5] introduce an ontological framework tailored for extracting information from diverse scientific sources.Theirframeworkintegratesmultipleontologiesto handleheterogeneousdata,demonstratingtheversatilityand scalabilityofOBIEinmultidisciplinaryscientificresearch.
Somodevilla García et al. [6] provide an overview of ontologylearningtasks,outliningtheprocessesinvolvedin automaticallygeneratingontologiesfromdata.Theirwork highlightstheadvancementsinmachinelearningandnatural language processing that support these tasks, which are fundamentalfordevelopingrobustOBIEsystems.
Krishnan et al. [7] explore semantic-based ontology mapping for mobile learning resources. They present a method that enhances information retrieval by mapping learning resources to ontologies, thereby improving the semantic understanding and relevance of retrieved information.
Islametal.[8]offeranoverviewoftheSemanticWeband introduce a .NET-based tool for knowledge extraction and ontology development. This tool demonstrates practical applications of OBIE technologies in industrial settings, highlighting the versatility of OBIE beyond academic research.
SharmaandKumarproposeanovelapproachtosemantic document indexing using machine learning and OBIE [9]. Theirmethodenhancesinformationretrievalbyintegrating machinelearningtechniqueswithontology-basedsemantic indexing,improvingtheprecisionofsearchresultsinlarge documentrepositories.
Etudoand Yoon presentanOBIE approachforlabeling radicalonlinecontentusingdistantsupervision[10].Their framework addresses the challenges of identifying and categorizingradicalcontentonline,demonstratingthesocial andethicalimplicationsofOBIEinmonitoringandmitigating onlineextremism.
Bashir and Warraich conduct a systematic literature review on the Semantic Web for distance learning, highlighting the role of OBIE in enhancing educational resourcesand delivery [11].Their review underscores the transformativepotentialofOBIEincreatingmoreinteractive andadaptivelearningenvironments.
Qiuetal.[12]exploretheapplicationofOBIEinmineral exploration, integrating text mining and deep learning methods. Their work exemplifies the use of OBIE in specializedscientificfields,showcasingitsabilitytoextract valuableinsightsfromcomplexgeologicaldata.
Hari and Kumar investigate ontology learning from unstructured text using transformers [13]. Their study highlightstheadvancementsinnaturallanguageprocessing that enable more effective extraction and structuring of informationfromlargetextcorpora.
Al-Aswadi et al. [14] focus on enhancing concept extractionforontologylearningusingdomaintimerelevance. Their research addresses the temporal dynamics of data, improving the accuracy of concept extraction in timesensitivedomains.
The multifaceted applications and methodologies of ontology-based information extraction across diverse scientificdomains.Fromenhancingacademicrepositoriesto monitoring online content and exploring geological data, OBIE proves to be a versatile and powerful tool for knowledge harvesting.Future research shouldcontinue to refine these techniques, addressing challenges such as scalability,interoperability,andreal-timedataprocessingto furtherexpandthecapabilitiesandapplicationsofOBIE.
3. METHODOLOGY
Ontology-BasedInformationExtraction(OBIE)employs diverse methodologies to harvest knowledge from unstructuredandstructureddataacrossscientificdomains. Thissectioncriticallyreviewsthemethodologiesadoptedin OBIE, as elucidated in the referenced studies, highlighting their contributions, techniques, and integration with advancedtechnologies.
3.1 Semantic Web Integration and Ontology Learning
Martinez-Rodriguez et al. [1] discuss the integration of information extraction (IE) with the Semantic Web, emphasizing the use of ontologies to enhance data interoperability and semantic richness. Their survey categorizesvariousIEmethodologiesthatleveragesemantic annotationsandlinkeddataprinciplestoimproveextraction accuracyandrelevance.Keytechniquesinclude:
1.Semantic Annotation: Linkingextractedentities andrelationshipstoexistingontologiestoenhance contextualunderstanding.
2.Linked Data: Utilizing RDF and SPARQL for querying and integrating heterogeneous data sources.
International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 11 Issue: 08| Aug2024 www.irjet.net p-ISSN:2395-0072
Petruccifocusesonlearningexpressiveontologies,whichare crucial for sophisticated OBIE systems [2]. The methodologiesinclude:
1.Ontology Enrichment: Enhancing existing ontologies with new concepts and relationships derivedfromtextcorpora.
2.Ontology Alignment: Harmonizing multiple ontologies to ensure consistency and interoperability.
3.2 Review of OBIE Techniques
SuganyaandPorkodiprovideacomprehensivereviewof OBIEmethodologies,outliningvarioustechniquesusedfor structureddataextraction[3].Theseinclude:
1.Rule-Based Approaches: Utilizing manually crafted rules to extract specific information from text.
2.Machine Learning Approaches: Employing supervisedandunsupervisedlearningalgorithmsto identifypatternsandextractinformation.
3.3 Frameworks for Specific Applications
Jose et al. [4] propose an OBIE framework tailored for academic knowledge repositories. Their methodology includes:
1.Entity Recognition: Identifying academic entities suchasauthors,publications,andinstitutions.
2.Relation Extraction: Determining relationships between entities, such as authorship and citation networks.
3.Integration with Academic Ontologies: Mapping extracted data to academic ontologies to ensure semanticcoherence.
Zamanetal.[5]introduceanontologicalframeworkfor diversescientificsources,highlightingmethodologiesthat addressheterogeneityandscalability:
1.Multi-Ontology Integration: Combining multiple domain-specific ontologiestohandlevariedscientific data.
2.SemanticInference: Applyingreasoningtechniques toinfernewknowledgefromextracteddata.
3.4 Ontology Learning Tasks
Somodevilla García et al. [6] outline ontology learning tasks essential for developing OBIE systems. Their methodologyincludes:
1.Concept Extraction: Identifyingkeyconceptsfrom text.
2.Relation Learning: Discovering relationships betweenconcepts.
3.Ontology Population: Automatically adding instancestotheontologybasedonextracteddata.
3.5 Semantic-Based Ontology Mapping
Krishnan et al. [7] explore semantic-based ontology mapping for mobile learning resources, employing methodologiessuchas:
1.Ontology Mapping: Aligning mobile learning resources with educational ontologies to enhance retrieval.
2.SemanticEnrichment: Addingsemanticmetadatato learningresourcesforimprovedcontextandrelevance.
3.6 Tools and Frameworks for Knowledge Extraction
Krishnan et al. [7] explore semantic-based ontology mapping for mobile learning resources, employing methodologiessuchas:
1.Ontology Mapping: Aligning mobile learning resources with educational ontologies to enhance retrieval.
2.SemanticEnrichment: Addingsemanticmetadatato learningresourcesforimprovedcontextandrelevance.
3.7 Machine Learning and OBIE
Krishnan Sharma and Kumar propose a novel approach combining machine learning with OBIE for semantic documentindexing[9].Theirmethodologyincludes:
1.Semantic Indexing: Using ontologies to index documentsbasedonsemanticcontent.
2.Machine Learning Integration: Applying machine learning algorithms to enhance the accuracy and efficiencyofindexing.
International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 11 Issue: 08| Aug2024 www.irjet.net p-ISSN:2395-0072
3.8 Distant Supervision for Labelling Content
EtudoandYoon presentanOBIEframeworkusing distantsupervisiontolabelradicalonlinecontent[10].Key methodologiesinclude:
1.DistantSupervision: Leveragingexternalknowledge basestogeneratetrainingdataforsupervisedlearning models.
2.Content Labelling: Automatically categorizing contentbasedonextractedsemanticfeatures.
3.9 Applications in Distance Learning and Geology
EtudoBashirand Warraich review theroleof the Semantic Web in distance learning, highlighting OBIE methodologiessuchas[11]:
1.Adaptive Learning Systems: Utilizing OBIE to personalize learning experiences based on extracted educationalcontent.
Qiu et al. [12] apply OBIE to mineral exploration data,integratingtextmininganddeeplearningmethods:
1.Text Mining: Extractinggeologicalinformationfrom unstructuredtexts.
2.DeepLearning: Applyingneuralnetworkstoimprove extractionaccuracyandhandlecomplexdatapatterns.
3.10 Transformer-Based Ontology Learning
HariandKumarinvestigatetheuseoftransformers for ontology learning from unstructured text, focusing on [13]:
1.TransformerModels: UtilizingadvancedNLPmodels toextractandstructureinformation.
2.Word Sense Disambiguation (WSD): Enhancing concept extraction by resolving ambiguities in word meanings.
3.11 Enhancing Concept Extraction
Hari Al-Aswadi et al.[14] focus on enhancing conceptextractionusingdomaintimerelevance,employing methodologiessuchas:
1.Temporal Analysis: Incorporating time-based relevancetoimprovetheaccuracyofconceptextraction.
2.Domain-Specific Customization: Tailoring extraction techniques to specific domains to enhance precision.
The methodologies reviewed demonstrate the versatility and sophistication of OBIE across various scientific domains. By integrating semantic technologies, machinelearning,anddomain-specificcustomization,OBIE methodologieseffectivelyextractandstructureknowledge, pavingthewayforadvancedapplicationsindiversefields. Future research should continue to explore innovative techniquesandtoolstofurtherenhancethecapabilitiesof OBIE.
4. COMPARATIVE ANALYSIS
Ontology-Based Information Extraction (OBIE) methodologies span a diverse array of techniques and approaches. This section compares these methodologies, summarizingtheirkeyfeatures,strengths,andlimitations, andprovidingacomparativetableforclarity.
4.1 Semantic Web Integration and Ontology Learning
Martinez-Rodriguez et al. [1] focus on the integrationofInformationExtraction(IE)withtheSemantic Web.Theirmethodologyutilizessemanticannotationsand linkeddataprinciplestoenhancedatainteroperabilityand semantic richness. While effective, this approach can be complexandrequiresadeepunderstandingofontological structuresandlinkeddatatechnologies.
Petrucciemphasizeslearningexpressiveontologies through ontology enrichment and alignment [2]. This approachensuresdetailedandaccurateontologycreation but can be computationally demanding and necessitates considerabledomainexpertise.
4.2 General OBIE Techniques
Suganya and Porkodi review various OBIE methodologies,includingrule-basedandmachinelearning approaches [3]. Rule-based methods are simple to implement but lack flexibility and scalability. Machine learning approaches, on the other hand, offer better adaptability and scalability but require extensive training dataandcomputationalresources.
4.3 Frameworks for Specific Applications
Jose et al. [4] propose a framework for academic knowledge repositories, focusing on entity recognition, relation extraction, and integration with academic ontologies.Thisframeworkishighlyeffectiveforacademic databutmaynotgeneralizewelltootherdomainswithout substantialmodifications.
Zaman et al. [5] introduce a framework for extractinginformationfromdiversescientificsources.Their approach integrates multiple ontologies and employs
International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 11 Issue: 08| Aug2024 www.irjet.net p-ISSN:2395-0072
semantic inference, making it versatile and scalable. However, the complexity of managing multiple ontologies canbeasignificantchallenge.
4 4 Ontology Learning Tasks and Semantic-Based Mapping
Somodevilla García et al. [6] discuss essential ontologylearningtasks,suchasconceptextraction,relation learning, and ontology population. Their structured approach is foundational for robust OBIE systems but requiresadvancedNLPandmachinelearningtechniques.
Krishnanetal.[7]focusonsemantic-basedontology mappingformobilelearningresources.Theirmethodology enhancesinformationretrievalthroughontologymapping andsemanticenrichment,providingimprovedcontextand relevancebutislargelyconfinedtoeducationalcontexts.
4.5 Tools for Knowledge Extraction and Machine Learning Integration
Islam et al. [8] present a .NET-based tool for knowledgeextractionandontologydevelopment,facilitating automatedontologyconstructionandknowledgeextraction. However, its capabilities may be limited by the .NET platformandintegrationchallengeswithothersystems.
SharmaandKumarintegratemachinelearningwith OBIEforsemanticdocumentindexing,enhancingprecision inlargedocumentrepositories[9].Thisapproachishighly effectivebutdemandssignificantcomputationalresources andadvancedmachinelearningexpertise.
4.6 Distant Supervision and Domain-Specific Applications
EtudoandYoonutilizedistantsupervisiontolabel radicalonlinecontent,leveragingexternalknowledgebases fortrainingdata[10].Thisapproachiseffectiveforcontent monitoring but may face challenges in maintaining up-todateandcomprehensiveknowledgebases.
Qiu et al. [12] apply OBIE to mineral exploration datausingtextmininganddeeplearning.Theirmethodology offershighaccuracyandhandlescomplexdatapatterns,but requires significant domain-specific customization and computationalresources.
4.7 Transformer-Based Learning and Temporal Relevance
HariandKumarinvestigatetheuseoftransformers for ontology learning from unstructured text, leveraging advancedNLPmodelstoextractandstructureinformation [13]. This approach is powerful but computationally intensiveandrequiressubstantialtrainingdata.
Al-Aswadi et al.[14] enhance concept extraction using domain time relevance, incorporating temporal analysistoimproveaccuracy.Thisapproachiseffectivefor dynamic domains but may require frequent updates to maintainrelevance.
Table -1: ComparativeTable
Methodology Key Techniques
Semantic Web Integration
Semantic Annotations ,Linked Data
Expressive Ontology Learning Ontology Enrichment, Alignment
Strengths
Highdata interoperabilit y,Semantic richness
Detailed, Accurate ontologies
Limitations
Complex implementati on,Requires deep ontological knowledge
Computationa llyintensive, Requires domain expertise
GeneralOBIE Techniques Rule-Based, Machine Learning
Academic Knowledge Framework Entity Recognition ,Relation Extraction
Diverse Scientific Framework
Ontology Learning Tasks
SemanticBased Mapping
Knowledge Extraction Tool
ML Integration forIndexing
MultiOntology Integration, Semantic Inference
Concept Extraction, Relation Learning
Ontology Mapping, Semantic Enrichment
Automated Ontology Constructio n
Machine Learning, Semantic Indexing
Simple implementatio n(RuleBased), Scalability (Machine Learning)
Effectivefor academicdata
Versatile, Scalable
Foundational forOBIE systems
Improved contextand relevance
Practical, Facilitates knowledge extraction
Highprecision
Lackof flexibility (Rule-Based), High computationa lresource requirement (Machine Learning)
Limited generalizabilit y
Complex management ofmultiple ontologies
Requires advancedNLP andML techniques
Largely confinedto educational contexts
Platform limitations, Integration challenges
Requires significant computationa lresources
International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 11 Issue: 08| Aug2024 www.irjet.net
Distant Supervision
DomainSpecific Applications
Distant Supervision, Content Labeling
Text Mining, Deep Learning
TransformerBased Learning Transforme rs,WSD
Temporal Relevance
Temporal Analysis
Effectivefor content monitoring
Highaccuracy, Handles complexdata patterns
Powerful extraction, Structured information
Improved accuracyfor dynamic domains
Maintaining up-to-date knowledge bases
Domainspecific customization , Computationa llyintensive
Computationa llyintensive, Requires substantial trainingdata
Frequent updates needed
The comparison reveals a diverse landscape of methodologies in OBIE, each with unique strengths and limitations. The choice of methodology depends on the specificrequirementsofthedomain,thecomplexityofthe data, and the available resources. Future research should focus on hybrid approaches that combine the strengths of different methodologies to address their individual limitations.
5. CONCLUSION
Thispaperhighlightedthesignificantadvancements and diverse applications of Ontology-Based Information Extraction (OBIE) across multiple scientific fields. The integrationofontologieswithsemanticwebtechnologieshas proven to be a robust method for enhancing data interoperabilityandsemanticrichness,asdemonstratedby Martinez-Rodriguez et al. [1] and others. This review systematically covered various methodologies such as ontology enrichment, alignment, machine learning-based approaches, and semantic-based ontology mapping. The methodologiesspanfromgeneralOBIEtechniquestospecific frameworks for academic repositories and scientific data extraction,showcasingtheversatilityandcapabilityofOBIE instructuringandextractingmeaningful informationfrom vastdatasets.
Thecomparativeanalysisunderscoresthestrengthsand limitationsofeachmethodology,revealingalandscapethat balances between simplicity, computational demands, and domain specificity. While methods like rule-based approachesoffereaseofimplementation,machinelearning and advanced NLP techniques like transformers present more scalable and precise solutions, albeit with higher computationalcosts.Thepapersuggeststhatfutureresearch
p-ISSN:2395-0072
shouldaimathybridmodelsthatleveragethestrengthsof variousOBIEtechniquestoovercomeindividuallimitations, focusingonenhancingscalability,real-timedataprocessing, andimprovedinteroperability.Thiscomprehensivereview not only provides a detailed synthesis of current OBIE methodologiesbutalsopavesthewayforinnovativeresearch directions to further enhance the field's capabilities and applications.
REFERENCES
[1] Martinez-Rodriguez, Jose L., Aidan Hogan, and Ivan Lopez-Arevalo. "Information extraction meets the semanticweb:asurvey."SemanticWeb11,no.2(2020): 255-335.
[2] Petrucci, Giulio. "Information extraction for learning expressive ontologies." InThe Semantic Web. Latest AdvancesandNewDomains:12thEuropeanSemantic WebConference,ESWC2015,Portoroz,Slovenia,May 31 June4,2015.Proceedings12,pp.740-750.Springer InternationalPublishing,2015.
[3] Suganya, G., and R. Porkodi. "Ontology based informationextraction-areview."In2018International Conference on Current Trends towards Converging Technologies(ICCTCT),pp.1-7.IEEE,2018.
[4] Jose, Veena, V. P. Jagathy Raj, and Shine K. George. "Ontology-basedinformationextractionframeworkfor academicknowledgerepository."InProceedingsofFifth International Congress on Information and Communication Technology: ICICT 2020, London, Volume2,pp.73-80.SpringerSingapore,2021.
[5] Zaman, Gohar, Hairulnizam Mahdin, Khalid Hussain, JemalAbawajy,andSalamaA.Mostafa."Anontological framework for information extraction from diverse scientificsources."IEEEaccess9(2021):42111-42124.
[6] Somodevilla García, María,Darnes Vilariño Ayala, and Ivo Pineda. "An overview of ontology learning tasks."ComputaciónySistemas22,no.1(2018):137146.
[7] Krishnan, Kalyani, Reshmy Krishnan, and Ayyakannu Muthumari. "A semantic-based ontology mapping–information retrieval for mobile learning resources."International Journal of Computers and Applications39,no.3(2017):169-178.
[8] Islam,Noman,DarakhshanSyed,andZubairA.Shaikh. "SemanticWeb:AnOverviewanda.net-basedToolfor Knowledge Extraction and Ontology Development."Semantic Technologies for Intelligent Industry4.0Applications:169-197.
International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 11 Issue: 08| Aug2024 www.irjet.net p-ISSN:2395-0072
[9] Sharma,Anil,andSureshKumar."Machinelearningand ontology-basednovelsemanticdocumentindexingfor information retrieval."Computers & Industrial Engineering176(2023):108940.
[10] Etudo, Ugochukwu, and Victoria Y. Yoon. "Ontologybasedinformationextractionforlabelingradicalonline contentusingdistantsupervision."InformationSystems Research35,no.1(2024):203-225.
[11] Bashir, Faiza, and Nosheen Fatima Warraich. "Systematic literature review of Semantic Web for distance learning."Interactive Learning Environments31,no.1(2023):527-543.
[12] Qiu,Qinjun,MiaoTian,LiufengTao,ZhongXie,andKai Ma. "Semantic information extraction and search of mineral exploration data using text mining and deep learning methods."Ore Geology Reviews(2024): 105863.
[13] Hari, Akshay, and Priyanka Kumar. "WSD based Ontology Learning from Unstructured Text using Transformer."ProcediaComputerScience218(2023): 367-374.
[14] Al-Aswadi,FatimaN.,HuahYongChan,andKengHoon Gan. "Enhancing relevant concepts extraction for ontology learning using domain time relevance."InformationProcessing&Management60, no.1(2023):103140.